【已解决】doris 2.0.4 自动收集统计信息oom

fe.warn.log中持续输出查询oom，为收集统计信息sql，如下

SELECT
	CONCAT('370740', '-', '-1', '-', 'time') AS `id`,
	0 AS `catalog_id`,
	370739 AS `db_id`,
	370740 AS `tbl_id`,
	-1 AS `idx_id`,
	'time' AS `col_id`,
	NULL AS `part_id`,
	54572517117 AS `row_count`,
	SUM(`t1`.`count`) * COUNT(1) / (SUM(`t1`.`count`) - SUM(IF(`t1`.`count` = 1, 1, 0)) + SUM(IF(`t1`.`count` = 1, 1, 0)) * SUM(`t1`.`count`) / 54572517117) as `ndv`,
	IFNULL(SUM(IF(`t1`.`column_key` IS NULL, `t1`.`count`, 0)), 0) * 11.279462585593992 as `null_count`,
	SUBSTRING(CAST('2021-04-13 20:31:40' AS STRING), 1, 1024) AS `min`,
	SUBSTRING(CAST('2024-05-27 10:40:11' AS STRING), 1, 1024) AS `max`,
	SUM(t1.count) * 16 * 11.279462585593992 AS `data_size`,
	NOW()
FROM
	(
	SELECT
		t0.`time` as `column_key`,
		COUNT(1) as `count`
	FROM
		(
		SELECT
			`time`
		FROM
			`internal`.`default_cluster:db`.`mytable` TABLET(3670690,
			3670746,
			3670790,
			3670834,
			3670882,
			4594011,
			3673494,
			.......,
			3673542,
			3673602,
			3673630) ) as `t0`
	GROUP BY
		`t0`.`time` ) as `t1`  

表mytable为大表(几百亿)，time为表dupkey的首字段，类型为datetime，TABLET()内约有百个，省略显示全部。百个tablets的数量肯定达到亿级，即使是group by后，所以肯定会oom。

问题：

1. 根据官网说明，默认所有表都是大表？huge_table_lower_bound_size_in_bytes默认值为0
2. 大表统计不是采样huge_table_default_sample_rows（400万）吗，上面的sql查一部分tablets是怎么来的
3. 关闭收集统计信息对性能有什么影响？
4. 能否只针对某些表关闭收集。